October 15, 2004
IBM has announced the latest in its eServer family based upon its converged POWER5 platform, formerly known as Squadrons. The new IBM eServer p5 595 and eServer i5 595 represent the flagship implementations of IBM's POWER5 server product line and Virtualization Engine technologies positioned as complete platforms that enable companies to meet their changing business and IT needs. The company also introduced the 32-way eServer p5 590 server. The new p5 systems deliver nearly three times the performance and twice the capacity of the pSeries 690 with the eServer p5 590 priced as much as 45% less. All of these new offerings are compatible with existing POWER4 processor-based servers and hence support the myriad existing ISV applications. The company noted that these new servers have the capacity to run more than 250 virtual servers on a single machine, which enables customers to consolidate data centers onto a single server while supporting hundreds of applications. The i5 version is focused on those who need an integrated business solution with multiple OS support (AIX 5L, i5/OS, Windows, and Linux), integrated IBM middleware, database, and security delivered in a scalable and flexible package. The eServer i5 595 delivers four times the performance of iSeries 890 32-way server. The eServer p5 590 is available with 1.65GHz clock speed and can scale to as many as thirty-two processors. The eServer p5 595 and eServer i5 595 are available with either 1.65GHz or 1.90GHz processors and can scale to a 64-way system. The new systems support AIX 5L Version 5.2 and 5.3, Linux, i5/OS, as well as Windows, on appropriately configured i5 systems. All of these POWER5 processor-based eServer systems will be available on November 19, 2004.
There are those who will be drawn to the sheer performance aspects of todayís announcement, and they should not be disappointed. However, as we are fond of saying, leadership leapfrogging by definition is an ongoing event. Nevertheless, these systems do offer a mind-numbing level of performance that not all that long ago would have been available only in mainframe class solutions. More important to our way of thinking is that the solutions are targeting different markets yet all share the same hardware platform. Differentiation is offered by configuration and software/middleware, high margin, and high value-add technologies. Hence, the value proposition of this platform is less affected by the ongoing competition for speeds and feeds supremacy that is taking place at the same time as falling hardware price points.
The fact that a single platform can effectively target such diverse constituencies as high-performance computing, commercial banking, multimedia imaging, SMB operations, and myriad third party applications is remarkable. While the p5 may be targeting those with UNIX workloads or extreme computing needs, the i5 offers an integrated, multi-workload solution that encompasses the vast majority of operating system environments found in the market today. The i5ís multi-workload ability is reconfirmed by the iSeries Virtualization Grand Slam benchmark, along with past benchmarks such as 3-in-1, which illustrate that high system utilization, multi-OS partitions, and virtualization are available to mere mortals without heavy investments in IT functionaries. Overall, we find these latest pSeries and iSeries offerings further testimony of the value of owning key intellectual property in systems design and taking advantage of the leverage that higher production volumes afford. IBM is taking an approach to the marketplace that protects past infrastructure investments by its customers while delivering increasing value in its new offerings. This is in sharp contrast to HP, who has effectively told its Alpha and PA-RISC customers to port their infrastructures to Itanium, or go pound sand. Thus, the UNIX, Linux, and integrated server communities will have a choice this November. One is to embark upon the path of the unknown future, typified by Itanium; the other is to join the standards path with a proven past and commitment to tomorrow, typified by POWER5.
Computer Associates (CA) has announced usage-based pricing, dubbed Measured Workload Pricing (MWP), for its mainframe management software products. MWP is an extension of the CA FlexSelect licensing program, and bases licensing costs on measured system usage on a pay-as-used basis, with enterprise IT able to increase usage as needed to meet growing demand. There is a baseline rate, with additional utilization charged according to a pre-determined on-demand rate schedule. This is in contrast to traditional licensing, which is based on hardware size and total processing capacity. Measured Workload Pricing uses quarterly reports generated by IBM's native Sub-Capacity Reporting Tool (SCRT) to tabulate and accrue utilization. Customers that have an existing CPU- or MIPS-based license can convert their existing license to a Measured Workload Pricing license. Available exclusively to MWP, the company is introducing twenty mainframe management solution sets, including multiple integrated products to support best practices. These solution sets span the disciplines of network, systems, applications, database, and storage management; job optimization; ILM; change and configuration management; identity and access management; testing and fault management; and infrastructure asset management and reporting. The company believes that the combination of usage-based pricing and best practices management capabilities will help it win back vulnerable mainframe business.
Several years ago, IBM began to change mainframe tool pricing aggressively. Although it had largely exited the tools business, IBM realized that mainframe customers were being overcharged for tools in comparison to other mainframe costs. The company also realized that without rationalized pricing, mainframes would not be competitive in the increasingly price-sensitive IT market. IBMís response was to make its own tools price competitive. This has forced other tool vendors such as CA to reassess how they price their tools or lose a lucrative business to IBM. CA has chosen usage-based licensing, which enables IT to maximize data center investments in several ways. IT pays for actual management software usage but not for more than is used. In our opinion the CA pricing and licensing model can be an effective tool to help enterprise IT contain costs and brings CA into alignment with IBMís Capacity On Demand (CoD) capabilities for zSeries. Usage-Based pricing also affords IT managers the flexibility to deploy applications with low usage, which might not be possible otherwise under the traditional pricing model. Of course, it also allows for system growth as well as reduction in scale.
CAís usage-based pricing and enhanced management tools are good for enterprise IT and necessary for CA. Businesses have more accurate usage models for data center resources, and are better able understand how new workloads affect overall TCO. From a vendor perspective usage-based pricing may be a mixed blessing. CA and other vendors will see lower revenues on enterprise essential core products, like system management, as pricing aligns more closely to usage. However, the alternative is to lose business to IBM or other software vendors adopting new licensing models. Vendors hope to offset such revenue declines by additional revenues from capacity changes requiring additional product. Once CA has its pricing aligned with other mainframe pricing, it will have to compete on a feature basis again. The mainframe market, like the rest of computing, has become increasingly cost-conscious in dynamic business environments. The days of large margins and locking customers into software without alternatives are over. The mainframe software market was once thought dead; long live the mainframe software market. May the best products win.
EMC has announced the acquisition of Dantz Development Corporation, a privately held eighty-employee supplier of backup and recovery software for SMBs, in an all cash deal valued at less than $50 million. Dantzís Retrospect products provide backup for file servers, desktops, notebooks, and business-critical applications, and offer such functionality as progressive incremental backup, automated desktop and notebook backup (PC & Mac), and media-independent one-step disk recovery. Dantz also brings with it a substantial customer base (500,000 registered users) including SMBs and operating units of larger enterprises, as well as established distribution channels, solutions partners, and patents. Dantz will continue to operate out of its Walnut Creek, California headquarters as part of the EMC software group.
Since 2000, EMC has been on acquisition frenzy, acquiring fifteen companies, including VMware, Documentum, and Legato. For EMC Dantz is a strategic acquisition whose entry-level products will complement EMC's disk-based backup-and-recovery product line and will enable EMC to quickly further extend its reach into the rapidly growing SMB market. In the last year, EMC has introduced the AX 100 SAN and NetWin110 NAS offerings for small businesses. The Dantz Retrospect offering will complement these products and represents EMC's first major SMB software entry. Dantz provides full backup, then incremental change backup, going forward. It is a lower-end disk-to-disk backup product without the richer Legato feature set and commensurate price tag. For example, Legato is a good solution for organization with the Clariion CX-series, but Dantz is a better fit for the AX100 product.
The Dantz acquisition is as much about channels and partners as it is about product and technology. Dantz currently sells approximately 25% of its products through OEMs such as Western Digital, Maxtor, Sony, and Iomega; another 15% to 20% through Web-based or telesales; and the rest through distributors such as Ingram Micro, Tech Data, and Bell Microproducts. These relationships will undoubtedly help EMC in its SMB quest. We believe that the Dantz acquisition is a good fit for EMC, but EMC still has challenges to overcome for its software to thrive in the SMB space. For most, Dantzís offerings are not as well recognized as competitors Veritas and CA; however, the EMC and Dantz combination is well positioned to achieve these objectives, as EMC brings the financial muscle for the heavy lifting and Dantz the requisite product and sales channels.
IBM announced this week it has acquired Montreal-based Systemcorp ALG Ltd, a company that offers project portfolio management software to enterprises. IBM said the new acquisition would be folded into its Rational Development suite, in order to give managers, executives, project managers, and other enterprise participants the ability to track multiple software development projects across the enterprise to determine the state of each project and whether or not more resources should be applied to any given project.
IBM said the new element of the Rational Software development environment would help enterprises align their IT resources deployment with the enterprises actual business goals. No terms of the acquisition were released.
Given the fact that a large percentage of IT projects end up over budget, underutilized, or poorly implemented, it makes good sense for any enterprise IT vendor to offer tools that help their customers manage the ratís nest of multiple company-wide IT initiatives. With this acquisition, IBM adds another interesting feature to its development environment that its sales people and partners can point to while pitching IBM offerings to prospective customers. This is not a revolutionary acquisition but its shows IBMís continued efforts to make their products compelling to an increasingly discerning customer base.
What is more interesting to us than this specific acquisition is the way in which smaller companies are being plucked out of obscurity to be folded into larger enterprises. Systemcorp is not just some start-up; it has a track record, customers, and revenues. Like many of the other smaller companies that have been scooped up by larger IT vendors in past years, it was already a going concern that had not gone the IPO route to fame and fortune (or bust). As the IT market continues to consolidate around larger vendors, the long-term fate of many smaller companies will be determined in large part by the fact they offer a better mousetrap to larger enterprises than their competitors. We believe that in the long run, such an environment will help speed innovations to a broader marketplace, as the scrutiny applied to these smaller companies by potential suitors will be much more stringent than the masses on Wall Street. In other words, wheat and chaff will fall appropriately in separate piles.
The United States Supreme Court decided not to take up a case involving Internet piracy and in doing so stopped an effort by the Recording Industry Association of America (RIAA) to force ISPs to hand over the names of customers who are using their Internet connection as a means to file-swap or share. The Supreme Court refused to entertain an appeal of a 2003 appeals court decision that declared that ISPs could not be required to hand over user identities under subpoena power. The decision not to take the case under consideration also implies that the RIAA may not use the Digital Millennium Copyright Act as a means for finding Internet file-swappers. The RIAA continues to file lawsuits against file-swappers, using the lawsuits to identify them. The case originated in an RIAA attempt to force Verizon Communications to turn over the names of file-swappers using its network.
While it is clear that much of the file-swapping activity involves illegal copying and trading of copyrighted material, the recording industry continues to fight an uphill battle against piracy conducted on the Internet. To date most technology restraints have failed to make an impact or have been circumvented by sophisticated pirates. Legal remedies have had some effect but the problem persists. The decision by the Supreme Court to uphold the lower courtís ruling that a subpoena is not sufficient to force ISPs to divulge their customersí identity takes an arrow out of the RIAAís quiver, short on ammunition to begin with.
Piracy of all sorts of digital content continues to be a major problem for producers of content, be it music, software, or images. Whole countries seem to be awash in bootleg copies of whatever content is hot that month. At the same time, content providers must realize that new delivery and cost models have been created by the expanding reach of the Internet and its related technologies. The success of iPods ó and more importantly iMusic ó tells us the recording industry has yet to come to grips with the fact that people want to create their own music collections bought in pieces, not as a pre-designed package such as a $16 CD. The RIAAís continued legal war on file-swapping tells us that the industry still has a long way to go to understand the full impacts and opportunities available via Internet technologies. As such, the RIAA seems to be swimming upstream, instead of allowing the current to do some of the work for them. Enterprise IT vendors can learn a lesson here, as they move into new markets and bring forth new products. Trying to bend the larger world to an internal vendor vision is a lot more difficult than moving with the existing flow of market momentum and capitalizing on how to make the most of where the river takes you.